Terminal Server, Codepages and Multicountry

Marije_Brummel

This entry is as much as informative and subject to question/discussion.

At one of my Dutch customers we are implementing two subsidiaries in Lithuania. These are simple offices that do accounting, sales and purchases with pretty much standard NAV which also the HQ does.

Since the LT localisation is not very exiting we decided to have them all in one database. I know this is not officially best practice from a technical and upgrade point of view, but as most of you know the advantages of having it in one system weigh more than the disadvantages.

The implementation goes very well as for training, setup etc. But today we ran into an unexpected issue. I think...

I already realised that we would run into the codepage issue as described in my blog.

http://dynamicsuser.net/blogs/mark_brum ... epage.aspx

I took this as a fact and disabled the flag, told the management and users what to expect. Everyone happy.

Until we went into the Terminal Server. We are running Terminal Server 2008 and for what we have found out (and I am not expert) we can only run TS in one codepage that is determined at startup. So we figured, what if we use the LT codepage? Would that effect Dutch users. Testing proved (we tought) that this would work as we don't have strange characters in NL and the LT ansi codepage has all dutch characters.

But then we got a call from order entry that tbe barcodes where not printing anymore. Investigation proved that the barcode font is also ANSI based and did not work on LT codepage.

Al and al I did not have a very good feeling about running NL on a LT codepage and we decided to create a new terminal server for the LT users since we have VMWare and a TS license is not that expensive.

Any feedback? Has anyone managed to setup TS with multiple codepages for different users?

mdPartnerNL

I have had the same problem with two customers and barcodes and the only way I solved it was setting the start/stop character of the barcode in a codeunit on the PC of the customer (regional settings) and then compiling it on that PC.

So:
in NL the start/stop chars are Î and Ï,
in UK the start/stop chars are Ä and Á
(It depends on the codepage)

If this is your problem I would use a different codeunit (created on that PC) depending on the country which returns the correct chars. And hoping someone with more knowledge on ANSI comes up with a more elegant solution.

FDickschat

At one of my customers (which you happen to know Mark) we are running NAV in DEU, NLD, ENU/G, ESP, FRA and now in CHS. All the european langauages run fine with 1 collation / all the same codepage on Citrix. For the chinese guys we had to setup new citrix servers anyway as this is double byte.

The barcodes do not create a problem as we are not using fonts but an OCX http://www.activebarcode.com/ocx/.
A CU creates an image of the barcode and imports it back into a BLOB which is then printed -> No issues with printer drivers or codepages. Very simple programming. Just add 3 lines of code to your report and a picture box control to the sections.

A compiled version of the ocx (an older, limited version) can be found in the downloads which can be used to do the same just with an external program http://www.mibuso.com/dlinfo.asp?FileID=106.

Marije_Brummel

Thanks Franky. Hope al is well in Hamburg...

I've thought about the ocx and I think we should implement this as it is more stable then using a font. (But we don't have much time, you know the story).

Nevertheless I think that running LT collation on a NL terminal server is not really a good idea and rather invest in a new server.

This is the feedback I got here and on my blog as being the "Best Practice"

* Using images for barcode
* New TS for each codepage

Thanks again.

kine

If you are using the free 3of9 barcode, it is not working with non-English codepages. I have modified version which could work if you are interested (I had same problem few years for Czech language).